Center for Advanced Intelligence Project

This talk will be held in a hybrid format, both in person at AIP Open Space of RIKEN AIP (Nihonbashi office) and online by Zoom. AIP Open Space: *only available to AIP researchers.

DATE & TIME
Feb 6, 2025: 11:00 am – 12:00 pm (JST)

TITLE
XFADS: nonlinear state-space identification from neural recordings

SPEAKER
Prof. Il Memming Park (CATNIP Lab: Computational and Theoretical Neural Information Processing Laboratory)

ABSTRACT
State-space graphical models and the variational autoencoder framework provide a principled apparatus for learning dynamical systems from data. State-of-the-art probabilistic approaches are often able to scale to large problems at the cost of flexibility of the variational posterior or expressivity of the dynamics model. However, those consolidations can be detrimental if the ultimate goal is to learn a generative model capable of explaining the spatiotemporal structure of the data and making accurate forecasts. We introduce a low-rank structured variational autoencoding framework for nonlinear Gaussian state-space graphical models capable of capturing dense covariance structures that are important for learning dynamical systems with predictive capabilities. Our inference algorithm exploits the covariance structures that arise naturally from sample based approximate Gaussian message passing and low-rank amortized posterior updates — effectively performing approximate variational smoothing with time complexity scaling linearly in the state dimensionality. In comparisons with other deep state-space model architectures our approach consistently demonstrates the ability to learn a more predictive generative model. Furthermore, when applied to neural physiological recordings, our approach is able to learn a dynamical system capable of forecasting population spiking and behavioral correlates from a small portion of single trials.

BIOGRAPHY
Il Memming Park (박일, 朴逸) (https://catniplab.github.io/) is a Group Leader of the Neural Dynamics (CATNIP) lab at the Champalimaud Centre for the Unknown. He was a tenured Associate Professor at Stony Brook University (2015-2023). He received his B.S. in computer science from KAIST, M.S. in electrical engineering and Ph.D. in biomedical engineering from the University of Florida (2010), and trained with Jonathan Pillow at University of Texas at Austin as a postdoctoral fellow (2010-2014). His research goal is to understand how information is represented and computations are implemented in artificial and biological neural systems over multiple time scales. He designs interpretable statistical models and machine learning methods specialized for neural signal processing.